The runaway mitigation coil papers are the four-paper modeling sequence published between 2021 and 2024 that works out whether a passive, non-axisymmetric in-vessel coil can stop 📝SPARC from generating a 📝runaway electrons beam during a 📝disruption.
The idea is that a 📝runaway electron mitigation coil needs no detection and no trigger. The disruption's own current-quench loop voltage drives current through the coil, whose three-dimensional field breaks the magnetic flux surfaces into stochasticity and lets fast electrons escape before the avalanche can multiply them. 📝Alex Tinguely and ten co-authors opened the sequence in Nuclear Fusion in 2021, chaining NIMROD 📝magnetohydrodynamics to ASCOT5 transport coefficients and the DREAM kinetic solver, and concluding that such a coil would completely prevent beam formation in SPARC.
The three papers after it are largely the sequence checking itself. V. A. Izzo led the 2022 study extending the modeling to 📝DIII-D as well as SPARC and deliberately hardening the assumptions — zeroing transport in regions where closed flux surfaces have re-healed — which found that only the n = 1 coil geometry of three considered has enough resonant content, and that under the harshest assumptions beams of 1 to 2 megaamps reappear. Tinguely's 2023 Plasma Physics and Controlled Fusion paper asks the inverse question, bounding the runaway current at roughly 1 megaamp against a pre-disruption 📝plasma current of 8.7, and identifying core-localized electrons inside r/a < 0.3 as the population that matters. Izzo's 2024 paper replaces the ideal wall with a resistive one and moves the simulation boundary out to the 📝vacuum vessel; on a DIII-D inner-wall-limited equilibrium the predicted runaway loss fraction rises from 90 percent to above 99.
No passive 3D coil has ever faced a disruption on any tokamak. Every number in these four papers is an output of simulation, and the sequence's own history — a result that weakened under conservative transport assumptions and then strengthened under better boundary conditions — is the evidence for how much the answer still depends on modeling choices. It is nonetheless the coil 📝Ryan Sweeney's 2026 📝ARC disruption strategy proposes for the power plant.
We are designing SPARC around a coil that has never been tested in a disruption. Publishing all four papers, including the one where our own conservative assumptions let a megaamp beam back in, is how that gets checked before the machine runs.